Consumers are far more willing to postpone spending today in order to have more in the
future than is generally believed, according to a study of 2,000 households by Arie Kapteyn
and Federica Teppa. Their findings, published in the March 2003 issue of the Economic
Journal, reveal that people typically have a strong preference for a steady path of increasing
consumption rather than one that rises sharply but may fall back. So rather than borrowing to
spend more today, on the whole, they prefer to have some more to spend tomorrow.
The researchers use data from an internet -based panel study in the Netherlands (the
CentERpanel, http://centerdata.uvt.nl/), which comprises around 2,000 households who
answer questions over the internet every weekend. Two sets of ‘stated preference’
experiments have been conducted with the households, one in 1998 and one in 2002. Each
time, people are asked to choose between a number of different consumption paths over an
extended period, typically several decades. The results show that households have a strong
preference for increasing consumption profiles: it is better to have some more to spend
tomorrow than today.
These results can be partly explained by ‘habit formation’: consumers get used to a certain
standard of consumption and compare today’s consumption to that standard. If consumption
goes up, consumption looks good compared with the standard, whereas if consumption goes
down, it looks bad. Thus, a gradually increasing consumption profile as one grows older is
preferred to a consumption profile that is flat or decreasing. But even for a given standard,
there is a preference for an upward sloping profile.
Economists have generally found it hard to tell plausible stories about how people make
trade-offs between expenditures in different periods. The standard central concept in their
models is the ‘time preference rate’ or ‘subjective discount rate’. Empirical studies typically
have found implausibly high values for individual subjective discount rates, in the order of
30% a year. Essentially this means that people are willing to take out a loan at an annual
interest rate of 30% in order to be able to consume today rather than a year from now.
Such outcomes are implausible, particularly considering long-term goals, like saving for
retirement. An increasing number of researchers therefore believe that consumers may have
several discount rates: one for the short term and one for the long term. This is known as the
‘hyperbolic discounting approach’.
Kapteyn and Teppa’s research concentrates on long-term decisions. In contrast to the rate of
30%, they find negative discount rates, in the order of minus 10%. This implies a strong
preference for postponing consumption: spending tomorrow has higher value than today.
Technically, this shows up as negative time preference rate: people would be willing to pay for
the privilege of postponing consumption until tomorrow.
The problem with most empirical studies is that time preference rates have to be inferred
indirectly, by observing the actual choices people make and by making assumptions about
the circumstances in which these choices are made. This research uses the ‘stated
preference’ approach, in which people make hypothetical choices. The advantage of this
approach is that the choices people make are completely observed and hence one does not
have to make many additional assumptions, except one: that people are serious when they
state which consumption path they prefer.
It is important to bear in mind that these experiments deal with hypothetical choices: this is
what people say they would ideally like to do. In practice, there are many temptations that
induce people to leave the virtuous path of abstinence today in order to have more tomorrow.
ENDS
Notes for Editors: ‘Hypothetical Intertemporal Consumption Choices’ by Arie Kapteyn and
Federica Teppa is published in the March 2003 issue of the Economic Journal.
Kapteyn is at RAND; Teppa is at CentER, Tilburg University.
The stated preference approach they use was pioneered by Robert Barsky, Tom Juster, Miles
Kimball and Matthew Shapiro of the University of Michigan in ‘Preference Parameters and
Behavioral Heterogeneity: an experimental approach in the Health and Retirement Study’,
published in the Quarterly Journal of Economics, Vol. 112, 1997, pp 729-58.
For Further Information: contact Arie Kapteyn on +1-310-393-0411 ext. 7973 (email:
Kapteyn@rand.org); or RES Media Consultant Romesh Vaitilingam on 0117-983-9770
or 07768-661095 (email: romesh@compuserve.com).